Bispecific T-cells expressing polyclonal repertoire of endogenous γδ T-cell receptors and introduced CD19-specific chimeric antigen receptor

Abstract

Even though other γδ T-cell subsets exhibit antitumor activity, adoptive transfer of γδ Tcells is currently limited to one subset (expressing Vγ9Vδ2 T-cell receptor (TCR)) due to dependence on aminobisphosphonates as the only clinically appealing reagent for propagating γδ T cells. Therefore, we developed an approach to propagate polyclonal γδ T cells and rendered them bispecific through expression of a CD19-specific chimeric antigen receptor (CAR). Peripheral blood mononuclear cells (PBMC) were electroporated with Sleeping Beauty (SB) transposon and transposase to enforce expression of CAR in multiple γδ T-cell subsets. CAR(+)γδ T cells were expanded on CD19(+) artificial antigen-presenting cells (aAPC), which resulted in >10(9) CAR(+)γδ T cells from <10(6) total cells. Digital multiplex assay detected TCR mRNA coding for Vδ1, Vδ2, and Vδ3 with Vγ2, Vγ7, Vγ8, Vγ9, and Vγ10 alleles. Polyclonal CAR(+)γδ T cells were functional when TCRγδ and CAR were stimulated and displayed enhanced killing of CD19(+) tumor cell lines compared with CAR(neg)γδ T cells. CD19(+) leukemia xenografts in mice were reduced with CAR(+)γδ T cells compared with control mice. Since CAR, SB, and aAPC have been adapted for human application, clinical trials can now focus on the therapeutic potential of polyclonal γδ T cells.

Figures

Figure 1

CAR+γδ T cells propagate on designer aAPC. (a) Transient (day 1) and stable (day 36) expression of CAR in T cells (top) and γδ T cells (bottom) in mock electroporated (“no DNA”) or CD19-specific CAR-electroporated cells (CD19RCD28). (b) Percentage of CAR+γδ T cells in the culture as transient (day 1) and stable (day 36) expression, where each shape represents an individual donor. (c) Rate of expansion of total γδ T cells (open triangles), CARnegγδ T cells (open squares), and CAR+γδ T cells (open circles) over tissue culture period following paramagnetic bead sorting (open arrow) and recursive stimulation (closed arrows) with aAPC and exogenous IL-2 and IL-21 administration. (d) Percentage-positive cells and mean fluorescence intensity of CD3, CAR, TCRαβ, and TCRγδ at day 36. Data are mean ± SD (n = 4) and quadrant percentages of flow plots are in upper right corner. aAPC, artificial antigen-presenting cell; ***P < 0.001. CAR, chimeric antigen receptor; IL, interleukin; TCR, T-cell receptor.

Figure 2

Immunophenotype of electroporated, separated, and propagated CAR+γδ T cells. (a) Expression by flow cytometry of cell surface markers associated with T cells and memory as gated on CD3+CAR+ cells. (b) Percentages of CAR+ T cells expressing T-cell markers, where each shape represents a different donor. Data are mean ± SD (n = 4). Quadrant percentages of flow plots are in upper right corner. CAR, chimeric antigen receptor; TCR, T-cell receptor.

Figure 3

Distribution of Vδ and Vγ in CAR+γδ T cells. (a) Representative FACS of Vδ populations (top) into Vδ1negVδ2neg (left), Vδ1+Vδ2neg (middle), and Vδ1negVδ2+ (right) populations and (b) Vδ allele mRNA expression in sorted T cells. (c) Vδ1negVδ2neg, Vδ1+Vδ2neg, and Vδ1negVδ2+ frequencies in gated CAR+γδ T-cell populations from four donors. (d) Vγ allele mRNA expression in CAR+γδ T cells. Data are mean ± SD (n = 3). Quadrant percentages of flow plots are in upper right corner. ***P <0.001. CAR, chimeric antigen receptor; FACS, fluorescence-activated cell sorting; TCR, T-cell receptor.

Figure 4

Bispecific γδ T cells produce proinflammatory cytokines when endogenous TCR and introduced CAR are stimulated. (a) CAR+γδ T cells at day 35 of co-culture on aAPC were stimulated for 4 hours with a mock cocktail (media alone) or leukocyte activation cocktail (LAC, PMA/ionomycin) to induce TCR stimulation and then analyzed by flow cytometry. CAR+ T cells were gated and tumor necrosis factor-α (TNF-α, top) and interferon-γ (IFN-γ, bottom) production is shown. (b) Luminex array (27-Plex) of cytokines secreted by CAR+γδ T cells in conditions described in a. (c) Similar to a except that EL4-CD19neg and EL4-CD19+ were used instead of mock/LAC. (d) Same as b but with EL4-CD19neg and EL4-CD19+ targets. Student's t-test for statistical analysis between mock and LAC (in b) and EL4-CD19neg and EL4-CD19+ (in d) where *P < 0.05, **P < 0.01, and ***P < 0.001. Data are representative of four donors for a and c and mean ± SD (n = 3) for b and d. aAPC, artificial antigen-presenting cell; CAR, chimeric antigen receptor; FGF, fibroblast growth factor; GM-CSF, granulocyte-macrophage colony-stimulating factor; IL, interleukin; MCP, monocyte chemoattractant protein; MIP, macrophage inflammatory protein; PDGFβ, platelet-derived growth factor-β PMA, phorbol 12-myristate 13-acetate; RANTES, regulated and normal T cell expressed and secreted; TCR, T-cell receptor; VEGF, vascular endothelial growth factor.

Figure 5

Specific lysis of CD19+ tumor cell lines by CAR+γδ T cells. (a) Standard 4-hour CRA of (a) CARnegγδ T cells against CD19+ B-ALL cell lines (REH, Kasumi-2, and Daudi-β2M) or primary CD19+ B cells from autologous (Auto) or allogeneic (Allo) donors, (b) CAR+γδ T cells against EL4-CD19neg (open squares) and EL4-CD19+ (closed squares) tumor cells, and (c) CARnegγδ T cells (open squares) and CAR+γδ T cells (closed squares) against CD19+ NALM-6 tumor cells. Data are mean ± SD from four healthy donors (average of triplicate measurements for each donor) that were pooled from two independent experiments. B-ALL, B-cell acute lymphoblastic leukemia; CAR, chimeric antigen receptor; CRA, chromium release assay; E:T, effector to target ratio.

Figure 6

In vivo antitumor activity of CAR+γδ T cells. (a) Schematic of experiment. (b) BLI derived from eGFP+ffLuc+CD19+ NALM-6 tumor and (c) representative images of mice at day 22. (d) Postmortem analysis of tissues and blood where tumor cells (CD19+eGFP+) were detected by flow cytometry. (e) Representative flow plots from d. Data are mean ± SD (n = 3–5 mice per group, representative of two independent experiments) and gating frequencies in e are displayed. The percentage of tumor cells is derived from detecting CD19+eGFP+NALM-6 by flow cytometry from postmortem samples. Statistics performed with (in b) two-way ANOVA with Bonferroni's post-tests and (in d) Student's t-test between treated and untreated mice. **P < 0.01 and ***P < 0.001. ANOVA, analysis of variance; BLI, bioluminescent imaging; CAR, chimeric antigen receptor; eGFP, enhanced green fluorescent protein; IL, interleukin; PBL, peripheral blood leukocyte.